Surrogate-based infill optimization applied to electromagnetic problems
Citations
197 citations
Cites background from "Surrogate-based infill optimization..."
...Of particular interest are the PoI and EI statistical criteria which are widely used for single-objective optimization [10,24]....
[...]
166 citations
Cites background or methods from "Surrogate-based infill optimization..."
..., optimization of a textile antenna (Couckuyt et al., 2010), identification of the elasticity of the middle-ear drum (Aernouts et al....
[...]
...The ooDACE Toolbox has already been applied successfully to a wide range of problems, e.g., optimization of a textile antenna (Couckuyt et al., 2010), identification of the elasticity of the middle-ear drum (Aernouts et al., 2010), etc....
[...]
144 citations
Cites background from "Surrogate-based infill optimization..."
...Whereas online data-driven EAs can use various infill sampling criteria [30], [31] to include additional training data for updating the surrogates during the optimization, offline data-driven EAs have no access to the real fitness evaluations and no model update can be carried out....
[...]
...A class of more formal model management strategies are known as infill sampling criteria [30], [31], which help select the next solution to be evaluated using the expensive fitness...
[...]
136 citations
References
11,357 citations
Additional excerpts
...[33]....
[...]
8,631 citations
6,914 citations
"Surrogate-based infill optimization..." refers methods in this paper
...[4, 27] as the Efficient Global Optimization (EGO) algorithm....
[...]
...This paper describes a popular optimization method for expensive black-box simulators based on kriging surrogate models, namely, expected improvement (EI) [4]....
[...]
...Therefore, the original EGO algorithm used kriging [4] as surrogate model of choice, since kriging provides analytical formulae for prediction as well as a point-wise error estimation....
[...]
6,583 citations
5,555 citations
"Surrogate-based infill optimization..." refers methods in this paper
...Actually, the co-kriging surrogate model [25, 26] is inherently a multifidelity surrogate model that essentially applies a correction to the output of the low-fidelity model....
[...]